[Laser] LED-based TRX

[C. Turner]

Hi Tim and the group,

Interesting results from Stuart:  It'll be interesting to see his 
comments firsthand on the list.  It would also be interesting to see 
pics, hear sounds and look at diagrams.

If a receiver is of the K3PGP-type (or of the KA7OEI Ver 3. type) or 
even VK7MJ's then I could easily believe that a "floating" detector 
diode could produce good output as even 1000pF cap in a 1 megohm circuit 
(that circuit is probably higher, so the lower "knee" frequency would 
likely be lower) would provide pretty good AC bypassing into the lower 
audio range.  Barring outside influences and too-small leakages that 
might prevent the charge on the FET's gate reaching equilibrium, it's 
essentially a capacitively-coupled (and probably fairly "light tolerant" 
receiver as constant DC biases are avoided) - except, of course, that 
the capacitor is placed on the "cold" side of the detector diode instead 
of the more-typical "hot" side (e.g. between the diode and the input 
device.)

***

A couple of years ago I was messing with building a simple T/R device 
using a single LED - in this case, a "Side-looking" Luxeon III in a 
parabolic-ish reflector - the idea being to to make a cheesy, 
short-range optical transceiver with all circuitry contained within the 
plastic housing:  The intent wasn't to build anything of that was 
practical or particularly high-performance - just to make something that 
"looked cool" and was capable of spanning only a couple kilometers at 
most as the small size and inaccuracy of the reflector yielded rather 
poor transmit beam divergence and receive focus.  One of the more 
difficult aspects of the design was to suspend a high-power LED out in 
the middle of space (at the focus of the reflector) and still dissipate 
several watts of heat:  You really can't put a heat sink out there as it 
tends to block too much light, so a copper rod was used for mounting the 
LED and the heat sink placed behind the reflector.

In testing as a detector I noted that with the Luxeon III it was easy to 
get hundreds of millivolts from the LED at ordinary room lighting and an 
extremely light load -  and it did, in fact, seem to be quite sensitive 
to modulated light in general at very low frequencies (10-100 Hz.)  One 
of the difficulties was, of course, that this high-power LED (like most 
high power LEDs) has a LOT of capacitance - some even get well even into 
the nanofarad range.  This murders high-frequency response and is even a 
bit too much to manage when a Transimpedance amp amplifier is used with 
all of the usual tricks (e.g. cascode, bootstrap, etc.) so I did some 
testing with back-biasing to make it more usable at "voice" frequencies, 
but the result was a *tremendous* amount of noise.

While I was at it, I did additional messing around with other, smaller 
(and lower capacitance) LEDs that I had on hand and was frankly 
surprised how well some of them worked:  The last time I'd tried this 
was years ago (80's-early 90's) using standard (dim) red, yellow and 
green devices - well before the "super high output" LEDs started 
appearing on the market - and while they worked, 
phototransistor/photodiode circuits worked far better.  While doing the 
more recent experiments it struck me that the newer breed of LEDs seem 
to be much "less bad" as detectors - but I've yet to place them on my 
"Photon range" and do any sort of scientific comparisons with more 
"traditional" detectors.  One thing in common with the various LEDs 
tested was that, in general, they do tend to get very noisy very quickly 
when back-biased at higher voltage - but that trait seems to vary wildly 
between different LEDs.  Again, this was just some informal testing, 
plucking disparate devices out of the LED bins and seeing what happened.

73,

Clint
KA7OEI